> Argument parsing is a deep topic and few libraries fit all use cases. I understand that. The my goal here is wanting to ‘bundle’ a collection of tiny, related applications (like the aforementioned ``git`` & ``pijul`` examples) but avoiding the startup times / executable size of dispatching to separate executables.
An alternative approach would be understanding how to best share code / an image if needed to accomplish said task instead of shipping multiple copies of very similar images. > Argument parsing is a deep topic and few libraries fit all use cases. Alternative 2 might be exposing more of the ``PRIVATE`` bits to consumers since many of the pieces are there to assemble your own system, but the exposed vocab kinda forces the consumer into using the vocab in a singular way. Unsure. > http://docopt.org I did see this mentioned in the origin argument parsing weblog post. But it does seem the project `would need to be resurrected if used <https://github.com/docopt/docopt/issues/519>`_. ------------------------------------------------------------------------ I think I should just let these ideas marinate in my mind me a while as there might a simpler solution — even if were some hack like throwing $INSERT_LANGUAGE’s argument parser to do the dispatch. But speaking of the `Re: Factor <https://re.factorcode.org>`_ blog… it is a great corpus of a decade of code examples & thought process — which is often quite humorous / entertaining. It’s a good read & valuable regardless of programming language background. _______________________________________________ Factor-talk mailing list Factor-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/factor-talk
